While adoptive immunotherapy holds promise as a treatment for malignant melanoma, development has been impeded by the lack of a reproducible and economical method for generating therapeutic numbers of antigen-specific CTL. Our preliminary data demonstrate that an artificial Antigen-Presenting Cell (aAPC), made by coupling HLA-Ig and anti-CD28 to beads, can reliably induce and expand antigen-specific CTL from healthy donors to a variety of different antigens. This raises the possibility that HLA-Ig can be used to replace standard dendritic cell based ex vivo expansion of antigen-specific CTL. Therefore the potential exists to use HLA-Ig-based aAPC as a viable method for induction, expansion and activation of clinical grade melanoma-specific T cells in the treatment of metastatic melanoma. In this study we propose to demonstrate functional efficacy of HLA-Ig based aAPC for inducing and expanding anti-tumor specific CTL from patients with metastatic melanoma. Specifically we propose to optimize aAPC stimulation of Mart-1 specific CTL by optimizing aAPC structure and the duration of stimulation. T cell activation requires delivery of a combination of signals through the T cell receptor (Signal 1) and through co-stimulatory molecules (Signal 2) such as engagement of CD28 by B7. The efficacy of the various formulations of aAPC, with variable ratios of signal 1 to signal 2 as well as different type of signal 2, will be determined. In vitro expanded Mart1-specific CTL will be studied for in vivo function in murine models, including a human/SCID model and possibly the murine A2-transgenic mice using adoptively transferred CTL. Efficacy of CTL treatments is likely to be augmented by transferring CTL populations directed at multiple antigenic epitopes. Therefore we will analyze the ability to induce/expand CTL specific for the melanoma associated, A2-restricted subdominant epitopes gpl00, NY-ESO-1 and tyrosinase. Effective immunotherapy will be augmented by also having melanoma-specific CD4 T cells. Our preliminary data indicates that using biotinylated class II-based aAPC or autologous DC, we can generate gpl00 melanoma specific CD4 cells. We propose to further develop the biotinylated class II HLA-based aAPC and to generate class II-lg-based aAPC for stimulation of melanoma specific class II-restricted CD4 T cells and analyze the importance of having both melanoma specific CD8 and CD4 T cells in the in vivo models. These studies will help evaluate the role of aAPC as a potential approach to adoptive immunotherapeutic for the treatment of metastatic melanoma.